In February 2026, the US Department of Labor published its AI Literacy Framework. It is the first national government framework to name prompting as a foundational skill. Not a nice-to-have. A foundational content area, alongside understanding AI principles and using AI responsibly.
The DOL framework is voluntary guidance, not a mandate. But it signals the direction of federal funding for workforce and education programmes in the United States, and it contains ideas that UK and international schools should pay attention to, regardless of whether they teach within the US system.
Five content areas in plain language
The DOL organises AI literacy into five content areas. Each one uses an imperative verb, which tells you something about the framework's philosophy. It is not about knowing things. It is about doing things.
Understand AI Principles. What AI is. How it works. What it can and cannot do. The difference between AI systems that follow rules and AI systems that learn from data. The limitations of AI, including that it produces confident outputs regardless of whether those outputs are accurate.
Explore AI Uses. Where AI is actually used, across industries and in daily life. Not theoretical applications. Real ones. AI in healthcare, logistics, finance, agriculture, creative industries, education. This content area asks students to look around and identify AI where they might not have noticed it.
Direct AI Effectively. This is the distinctive one. The DOL calls prompting a foundational skill. Writing clear instructions for AI systems. Providing context. Iterating when the output is not right. Knowing how to get useful results from AI by communicating clearly with it. No other national framework names this as a standalone content area.
Evaluate AI Outputs. Assessing whether what AI produces is accurate, complete, biased, or fit for purpose. This is the same competence that appears in every framework, from UNESCO to PISA to Anthropic's Discernment dimension. The DOL gives it its own content area because the skill is so fundamental.
Use AI Responsibly. Data privacy. Workplace policies. Accountability for AI-assisted decisions. Understanding when AI use is appropriate and when it is not. The ethical and practical dimensions of being a responsible AI user.
What the DOL gets right that other frameworks miss
The DOL framework is the most practical of the major frameworks. Where UNESCO provides principles and the OECD provides competences, the DOL provides content areas that sound like things a teacher could actually plan a lesson around.
"Direct AI Effectively" is the clearest example. Every teacher who has watched a student type a vague prompt into ChatGPT and accept whatever comes back knows this skill matters. The student who types "tell me about the Cold War" and gets a generic paragraph has not directed AI effectively. The student who types "explain three ways the Berlin Wall's construction in 1961 changed daily life for families in East Berlin, using specific examples" and then evaluates the response against their textbook has directed AI effectively and evaluated the output. Two content areas in one task.
No other framework names this as its own domain. The OECD folds it into "Create with AI." UNESCO folds it into "AI Techniques and Applications." Anthropic calls it "Description." The DOL pulls it out and says: this is important enough to stand alone. For teachers, that clarity is useful.
The DOL also comes with seven delivery principles that describe how AI literacy should be taught. Learning should be relevant to the student's context. It should be hands-on. It should build on what people already know. It should be inclusive and accessible. These principles are not unique, but the DOL states them as obligations on the provider of training, not just aspirations.
How the five content areas map to other frameworks
The DOL does not exist in isolation. Its five content areas map cleanly to the other frameworks schools are likely to be tracking.
DOL's "Evaluate AI Outputs" is the same competence as PISA 2029's "Evaluating Credibility," OECD AILit's "Engage with AI," and Anthropic's "Discernment." Five frameworks, five names, one skill.
DOL's "Use AI Responsibly" maps to UNESCO's "Ethics of AI," OECD AILit's "Manage AI," and Anthropic's "Diligence." The responsible use dimension appears everywhere because every framework agrees it is non-negotiable.
DOL's "Understand AI Principles" maps to UNESCO's "AI Techniques and Applications" and OECD AILit's "Engage with AI" at the foundational level. Understanding what AI is and how it works is the baseline that everything else builds on.
DOL's "Direct AI Effectively" maps most closely to OECD AILit's "Create with AI" and Anthropic's "Description." But the DOL gives it more weight than either of those frameworks.
DOL's "Explore AI Uses" maps to UNESCO's cross-cutting approach and OECD AILit's scenarios. It is the "look around you" dimension that grounds AI literacy in the real world.
AILitKit maps activities to all of these frameworks simultaneously. When a student practises evaluating an AI-generated paragraph in English, the framework alignment section tags it against DOL (Evaluate AI Outputs), PISA (Evaluating Credibility), OECD AILit (Engage with AI), Anthropic (Discernment), and UNESCO (AI Techniques and Applications). The teacher does not need to track this manually. The guide does it.
Why UK schools should care about a US framework
The DOL framework is voluntary US guidance. It has no regulatory force in the UK. So why should a UK school pay attention?
Three reasons.
First, it validates what you are already doing. If your school teaches students to evaluate information, question sources, and think critically about technology, the DOL framework tells you those are exactly the right skills. It just names them in language that is clearer than most education frameworks manage.
Second, it adds prompting as a named skill. UK schools are going to have to teach students how to communicate effectively with AI systems. The national curriculum review will address this. The DfE will eventually provide guidance. The DOL has done it first, and the language they use is a useful reference point.
Third, for international schools, the DOL framework is increasingly relevant to parents and employers. A British school in Dubai whose students will enter a US, European, or Gulf workforce benefits from demonstrating alignment with multiple frameworks. The DOL is the one that speaks the language of employment.
For our full comparison of how the DOL sits alongside UNESCO, OECD, PISA, and Anthropic, see our guide to AI literacy frameworks explained.
The delivery principles matter as much as the content areas
The DOL framework includes seven principles for how AI literacy should be delivered. These are easy to overlook because the content areas are more attention-grabbing. But the principles contain something important: the DOL does not define progression levels.
Unlike the OECD framework, which suggests primary and secondary scenarios, or the UAE MoE curriculum, which defines cycle-by-cycle progression, the DOL explicitly leaves progression to the provider. This is deliberate. The DOL says: here is what people need to learn. How you sequence it, how you scaffold it, and how you assess it is your decision.
For schools, this is both freedom and responsibility. The DOL gives you the destination but not the route. AILitKit's Introduced, Practised, and Assessed depth labels are our overlay on the DOL's content areas, providing the progression structure that the framework intentionally does not define. This is flagged in the system as a pedagogical scaffold, not a DOL-defined structure.
Where to start
If you are a UK school considering the DOL framework, do not add it as another tracking requirement. Use it as a lens.
Look at one lesson you are teaching this week. Ask which of the five content areas it could touch. Could students understand an AI principle? Explore how AI is used in this field? Direct an AI system to complete a subject-relevant task? Evaluate what it produces? Consider the responsible use implications?
Most lessons will touch two or three of the five. That is not a coincidence. The content areas overlap with skills teachers already develop. The framework just names them.
AILitKit maps activities to the DOL framework alongside UNESCO, OECD AILit, PISA 2029, and Anthropic 4D. See where your lessons connect. Try a free guide at ailitkit.com